cs448f: image processing for photography and vision

50
CS448f: Image Processing For Photography and Vision Blending and Pyramids

Upload: vangie

Post on 25-Feb-2016

50 views

Category:

Documents


6 download

DESCRIPTION

CS448f: Image Processing For Photography and Vision. Blending and Pyramids. Blending. We’ve aligned our images. What now? Averaging Weighted averaging min/max/median. Noise reduction by Averaging. 2 Shots. 4 Shots. 8 Shots. 16 Shots. Noise Reduction by Averaging. - PowerPoint PPT Presentation

TRANSCRIPT

Page 1: CS448f: Image Processing For Photography and Vision

CS448f: Image Processing For Photography and Vision

Blending and Pyramids

Page 2: CS448f: Image Processing For Photography and Vision

Blending

• We’ve aligned our images. What now?• Averaging• Weighted averaging• min/max/median

Page 3: CS448f: Image Processing For Photography and Vision

Noise reduction by Averaging

Page 4: CS448f: Image Processing For Photography and Vision

2 Shots

Page 5: CS448f: Image Processing For Photography and Vision

4 Shots

Page 6: CS448f: Image Processing For Photography and Vision

8 Shots

Page 7: CS448f: Image Processing For Photography and Vision

16 Shots

Page 8: CS448f: Image Processing For Photography and Vision

Noise Reduction by Averaging

• We’re averaging random variables X and Y• Both have variance S2

• Variance of X+Y = 2S2

• Std.Dev. of X+Y = sqrt(2) . S• Std.Dev of (X+Y)/2 = sqrt(2)/2 . S• Ie, every time we take twice as many photos,

we reduce noise by sqrt(2)

Page 9: CS448f: Image Processing For Photography and Vision

Noise Reduction by Averaging

• Average 4 photos: noise gets reduced 2x• Average 8 photos: noise gets reduced 3x• Average 16 photos: noise gets reduced 4x

Page 10: CS448f: Image Processing For Photography and Vision

Noise Reduction by Median

• (demo)

Page 11: CS448f: Image Processing For Photography and Vision

Median v Average

Page 12: CS448f: Image Processing For Photography and Vision

Median v Average

Page 13: CS448f: Image Processing For Photography and Vision

Can we identify the bad pixels?

• They’re unlike their neighbours• Instead of averaging, weighted average – where weight = similarity to neighbours

Page 14: CS448f: Image Processing For Photography and Vision

Weighted Average

Page 15: CS448f: Image Processing For Photography and Vision

Can we identify the bad pixels?

• They’re unlike their neighbours• Instead of averaging, weighted average – where weight = similarity to neighbours

• Favors blurriness

Page 16: CS448f: Image Processing For Photography and Vision

Input

Page 17: CS448f: Image Processing For Photography and Vision

Other uses of Median

• Removing Transient Occluders• (live demo)• (Gates demo)• (surf demo)

Page 18: CS448f: Image Processing For Photography and Vision

Panorama Stitching

Page 19: CS448f: Image Processing For Photography and Vision

Panorama Stitching

Page 20: CS448f: Image Processing For Photography and Vision

Panorama Stitching

Page 21: CS448f: Image Processing For Photography and Vision

Panorama Stitching

Page 22: CS448f: Image Processing For Photography and Vision

Panorama Stitching

Page 23: CS448f: Image Processing For Photography and Vision

Multiple Exposure Fusion

Page 24: CS448f: Image Processing For Photography and Vision

Multiple Exposure Fusion

Page 25: CS448f: Image Processing For Photography and Vision

Multiple Exposure Fusion

Page 26: CS448f: Image Processing For Photography and Vision

Multiple Exposure Fusion

Page 27: CS448f: Image Processing For Photography and Vision

Multiple Exposure Fusion

Page 28: CS448f: Image Processing For Photography and Vision

Multiple Exposure Fusion

Page 29: CS448f: Image Processing For Photography and Vision

Multiple Exposure Fusion

Page 30: CS448f: Image Processing For Photography and Vision

Multiple Exposure Fusion

Page 31: CS448f: Image Processing For Photography and Vision

Multiple Exposure Fusion

Page 32: CS448f: Image Processing For Photography and Vision

Multiple Exposure Fusion

Page 33: CS448f: Image Processing For Photography and Vision

Multiple Exposure Fusion

Page 34: CS448f: Image Processing For Photography and Vision

Focus Fusion

Page 35: CS448f: Image Processing For Photography and Vision

Focus Fusion

Page 36: CS448f: Image Processing For Photography and Vision

Focus Fusion

Page 37: CS448f: Image Processing For Photography and Vision

Focus Fusion

Page 38: CS448f: Image Processing For Photography and Vision

Focus Fusion

Page 39: CS448f: Image Processing For Photography and Vision

Pyramids

• We’ve been breaking images into two terms for a variety of apps– Coarse + Fine

• More generally we can break it into many terms:– Very coarse + finer + finer ... + finest.

Page 40: CS448f: Image Processing For Photography and Vision

Pyramids• We can do this by blurring more and more:

Page 41: CS448f: Image Processing For Photography and Vision

Pyramids• And then (optionally) taking differences

--

Page 42: CS448f: Image Processing For Photography and Vision

Pyramids• The coarse layers can be stored at low res.

GaussianPyramid

LaplacianPyramid

Page 43: CS448f: Image Processing For Photography and Vision

Pyramids• How much memory does this use?

Page 44: CS448f: Image Processing For Photography and Vision

Pyramid Uses:

• Sampling arbitrarily sized Gaussians

• Equalizing an image– The different levels represent different frequency ranges– We can scale each frequency level and recombine

• Blending multiple images

Page 45: CS448f: Image Processing For Photography and Vision

Pyramid Blending

• Key Insight:– Coarse structure should blend very slowly

between images (lots of feathering), while fine details should transition more quickly.

• More robust to tricky cases than plain old compositing

Page 46: CS448f: Image Processing For Photography and Vision

Inputs:

Page 47: CS448f: Image Processing For Photography and Vision

Compositing: Hard Mask

Page 48: CS448f: Image Processing For Photography and Vision

Compositing: Soft Mask

Page 49: CS448f: Image Processing For Photography and Vision

Multi-Band Blending

Page 50: CS448f: Image Processing For Photography and Vision

Exposure Fusion• http://research.edm.uhasselt.be/~tmertens/papers/exposure_fusion_reduced.pdf